MEMBRANE TECHNOLOGY: A METHOD OF GAS …

0

MEMBRANE TECHNOLOGY: A METHOD OF GAS SEPARATION

by

Skylar Addicks

Submitted in partial fulfillment of the

requirements for Departmental Honors in

the Department of Engineering

Texas Christian University

Fort Worth, TX

May 4, 2015

ii

MEMBRANE TECHNOLOGY: A METHOD OF GAS SEPARATION

Project Approved:

Supervising Professor: Walt Williamson, Ph.D.

Department of Engineering

Robert Bittle, Ph.D.

Department of Engineering

Kim Janzen

Outside Consultant

iii

ABSTRACT

In the Oil and Gas Industry many pieces of equipment in the production field are

fueled by the natural gas that is either produced on a well pad facility site or from a

nearby gas pipeline. The problem with using directly produced natural gas as fuel is that,

typically, it is very rich, or abundant in heavier hydrocarbons. The leaner the fuel,

meaning fuel consisting of lighter hydrocarbons such as methane and ethane, burn better

and have several benefits which are discussed later in this report. There are various

methods of conditioning fuel which aim to reduce the fuel’s energy content and increase

its molecular percent in methane and ethane. This allows for the engine to burn off a

cleaner fuel. In this particular test, the MTR FuelSep® system will be used as the method

of fuel conditioning. The MTR FuelSep® system functions with the use of membranes,

and this experiment was designed to validate its function, characterize its performance

map, determine if it is economical to use, and discover any other benefits such as

emission reduction. After gathering over 44 hours of testing data, a performance map was

created which associated the differential pressure across the membrane with its

performance index as well as its rate of return. Additionally, emission testing was

conducted which shows the MTR FuelSep® system significantly reduces the content of

CO and NOx in the compressor engine exhaust. Throughout this report more information

will be provided on membrane technology, the design procedures, the results of the

testing, and discussion of the unit’s performance as well as performance in the future.

iv

TABLE OF CONTENTS

INTRODUCTION .............................................................................................................. 1

Membrane Technology Background Information ....................................................... 1

Other Membrane Designs .................................................................................. 2

Fuel Gas Conditioning through FuelSep® ...................................................... 3

Other Applications of Membrane Technology ................................................ 4

Design of MTR FuelSep® Unit ..................................................................................... 4

Process Description ............................................................................................. 5

Purpose of Performance Testing .................................................................................... 7

METHODS AND MATERIALS ........................................................................................ 7

Manufacturer System Specification ............................................................................... 7

Manufacturer Process Connections................................................................................ 8

Manufacturer General Rules ........................................................................................... 9

Permeate Over-Pressure ..................................................................................... 9

Liquid Contact ................................................................................................... 10

Pressure Shock and Flow Surges .................................................................... 10

Maximum Differential Pressure Feed to Reside ........................................... 10

Maximum Operating Temperature ................................................................. 10

Oil Vapor ............................................................................................................ 10

HCL ..................................................................................................................... 11

Instrumentation Guide ................................................................................................... 11

PROCEDURES................................................................................................................. 12

Preparation for Start-Up Procedure ............................................................................. 12

v

Start Up Procedure of FuelSep® System .................................................................... 13

Skid Purge Procedure .................................................................................................... 14

Shut Down Procedure of FuelSep® System ............................................................... 14

Phase I: Testing for Function of FuelSep® System .................................................. 15

Phase II: Optimizing Performance Test of FuelSep® System ................................. 18

Phase III: Testing for Upsets of FuelSep® System ................................................... 19

Phase IV: Emissions Testing ........................................................................................ 21

RESULTS ......................................................................................................................... 22

Performance Index ......................................................................................................... 25

DISCUSSION ................................................................................................................... 26

Assumptions Made ......................................................................................................... 27

Economics ....................................................................................................................... 28

Qualitative Benefits ....................................................................................................... 29

Plans for Long Term Installation ................................................................................. 29

CONCLUSION ................................................................................................................. 30

REFERENCES ................................................................................................................. 31

vi

LIST OF TABLES

TABLE 1: Description of Major Parts ................................................................................ 8

TABLE 2: Instruments Used ............................................................................................ 11

TABLE 3: Operating Conditions at Well-Pad Facility ..................................................... 16

TABLE 4: Heat and Material Balance (Model Prediction) .............................................. 16

TABLE 5: Membrane Optimization Data ......................................................................... 22

TABLE 6: Emissions Testing Data.................................................................................. 24

vii

LIST OF FIGURES

FIGURE 1: Microscopic View of Membrane Layers ......................................................... 1

FIGURE 3: Hollow Fiber Module Cross Section (on left) and Side View (on right) ........ 3

FIGURE 4: Outline of MTR FuelSep® System ................................................................. 5

FIGURE 5: Flow Diagram of Membrane Unit at Well Pad Facility .................................. 6

FIGURE 6: Model of Membrane Performance at Well Pad Facility................................ 17

FIGURE 7: Nitrogen Oxide Emissions Reduction ........................................................... 24

FIGURE 8: Carbon Monoxide Emissions Reduction ....................................................... 24

FIGURE 9: Oxygen Emissions Reduction ....................................................................... 25

FIGURE 10: Internal Rate of Return and Performance versus Performance Index ......... 26

1

INTRODUCTION

Membrane Technology Background Information

The key to the FuelSep® recovery process is an organic-selective composite

membrane that permeates condensable vapors, such as C3+ hydrocarbons, aromatics, and

water vapor, while rejecting non-condensable gases, such as methane, ethane, nitrogen,

and hydrogen. MTR can achieve efficient separations by exploiting differences in

permeability through MTR’s robust, high-flux polymeric membrane. This is a relatively

new technology, first commercialized in 1990.

The membrane consists of a very thin, highly selective, rubbery top layer and a

tough, relatively open micro-porous support layer (see Fig.1). The top layer performs the

separation; the porous support layer provides mechanical strength. A non-woven fabric

serves as the backing material for membrane structure.

Figure 1: Microscopic View of Membrane Layers

For use in FuelSep® Processes, MTR incorporates membrane into spiral-wound

membrane modules. These modules contain membrane envelopes wound around a central

collection pipe. Mesh spacer materials create channels through which the feed gas and

permeate vapors travel. As a feed gas stream containing organic vapor passes across the

membrane surface, the organic fraction passes preferentially though the membrane and

2

enters the permeate channel. The permeate vapor spirals inward through the permeate

channel to the central collection pipe.

To provide the driving force for permeation, a pressure difference is maintained

across the membrane between the feed and the permeate stream. The pressure difference

is obtained by compressing the feed stream and having two back pressure regulators on

the residue and permeate outlet streams. The pressure difference directly affects the rate

at which nitrogen and organic vapor permeate the membrane. The larger the pressure

difference the greater the flux through the membrane, resulting in a reduction in the

number of membrane modules needed to perform a desired separation.

The membrane modules consist of densely packed sandwich of membrane and

spacers in a spiral wound configuration around a central permeate pipe (observe Fig. 2).

Figure 2: Side View of Membrane Function

Other Membrane Designs

Another type of membrane that could also perform fuel gas conditioning is the

hollow fiber module consisting of rigid, glassy polymers. (Shown in Figure 3 below). For

all membrane separations, the driving force is the differential pressure between the inlet

feed and the permeate side of the membrane. In rigid, glassy polymers the dominant

factor determining membrane selectivity is the ratio of the gas diffusion coefficients,

3

which is highly dependent on molecular size. Thus, glassy polymer membranes typically

permeate the smaller molecules, methane and ethane, and reject the larger molecules,

propane, butane, and higher hydrocarbons. For the rubbery polymer membranes like the

one being tested, the dominant factor for membrane diffusion is the ratio of the gas

solubilities, which reflects the ratio of the condensability of the components making it a

reverse selective behavior.

Figure 3: Hollow Fiber Module Cross Section (on left) and Side View (on right)

Fuel Gas Conditioning through FuelSep®

Natural gas is commonly used in the field as a fuel source for gas engines and

turbines in the gas processing industry. Frequently, raw natural gas is the only fuel

available to operate compressor stations with in remote locations and on offshore

platforms. The gas has a high heating value, high hydrocarbon dew point, and low octane

number, which can cause operating problems. In gas engines, the rich fuel may pre-

detonate which can severely damage the internals of the firing chamber. Also, the

condensation of hydrocarbons (due to day-night temperature variations) may damage

combustion chambers in gas engines and gas turbines, increasing maintenance cost and

downtime. Since the engines drive and turbines drive other machinery, any disruption in

their operation will reduce production resulting in significant revenue loss. To increase

4

the reliability and reduce unscheduled downtime of such key equipment, fuel

conditioning through membrane technology is a simple way to solve this problem. The

conditioned fuel gas is significantly depleted in the higher hydrocarbons, is completely

dehydrated, and still retains the heating value needed to drive a compressor.

Other Applications of Membrane Technology

Membrane technology is an umbrella term referring to the mechanical separation

processes of gaseous or liquid streams with the use of membranes. Although it is a

relatively young technology, it proves to have many benefits to a variety of industries.

Membrane systems can be found in water treatment using reverse osmosis, in waste water

purification because of ultra and microfiltration that can be achieved with membranes,

filtration in the food industry, for medical applications such as artificial kidneys and

lungs, as fuel cells, in gas separation processes, and these are just a few examples. One of

the reasons this technology is becoming more useful is because membranes use less

energy than other separation processes, and typically have no moving parts.

Design of MTR FuelSep® Unit

The MTR FuelSep® system sits on a skid that is approximately 5 ft. wide, 7 ft.

tall, and 23 ft. long, seen in Figure 4. No operator attention is required and since the

system has no moving parts, maintenance expenses are minimal. The expected membrane

life is from 3 to 5 years. There are three main parts to the makeup of this unit. The

separator, the filter coalescer, and the membrane modules. The filter coalescer is located

in the region farthest left in Figure 4. The separator is to the right of the filter coalescer,

and the membrane modules are housed in the longest portion of the membrane on the

5

right side of the system. The residue exits out the bottom of the unit while the permeate

exits out the end.

Figure 4: Outline of MTR FuelSep® System

Process Description

The purpose of this membrane is to generate a conditioned fuel gas, as well as

recover hydrocarbons from the nitrogen/hydrocarbon gas stream in order to provide

lower energy content gas streams for fuel gas.

The feed gas mixture enters the membrane vessels through the inlet ports and then

enters the feed spacer material in the membrane modules as it flow through the feed

spacer material, the gas mixture is in contact with the ultrathin separation layer. Part of

the H2S and hydrocarbons permeate the separation layer and flow via the micro-porous

support layer and the woven material, to the permeate spacer material. The amount of gas

that permeates is determined by the flowrate or residence time in the feed spacer material

in the membrane modules. This is controlled by the pressure on the permeate line. The

6

permeate gas flows through the spacer material to a central permeate collection pipe of

each membrane module. This gas then flows out the permeate connection to the permeate

piping.

The non-permeated gas- with H2S and heavy hydrocarbon portions mostly

removed at this point is collected as residue from the membrane module insert and flow

out of the membrane module housing through the residue nozzle to the residue piping

where it is sent to the fuel gas line.

The membrane module inserts are fitted with U-cup seals to avoid gas flow

between the membrane module insert and the membrane module pressure vessel, which

would allow a bypass around the membrane module insert. The membrane module

pressure vessels are closed with a flange on each end, one of which has a permeate outlet

port in the center of it.

Figure 5: Flow Diagram of Membrane Unit at Well Pad Facility

The inlet to the membrane is taken as a slip-stream from the HP discharge at

approximately 700 psig and is throttled to a significantly lower pressure (~430 psig)

before entering the membrane. The inlet gas first enters a 2-stage integrated filter

7

coalescer which removes any liquid condensates/aerosols formed and the clean gas enters

the membrane vessels. The membrane vessels split the inlet into two streams:

1. The membranes preferentially permeate the heavy hydrocarbons, and the

resulting permeate stream which is enriched in the heavy hydrocarbons, will

be recycled back to the suction header of the compressor at 85 to 100 psig.

2. The residue gas stream will be conditioned gas with a lower LHV of less than

1000 BTU/SCF. The residue gas will be throttled further down in pressure and

routed to the fuel header.

Purpose of Performance Testing

The goal of the MTR FuelSep® is to improve the quality of the fuel gas by

removing the heavy hydrocarbon content. Therefore, the purpose of this testing is to

determine whether the MTR unit performs as specified, how sensitive is the unit to

performance parameters, and what is the optimal operating conditions of this unit.

METHODS AND MATERIALS

Manufacturer System Specification

The membrane module unit consists of one 8” membrane module pressure vessel

and the 8” membrane module inserts. The module pressure vessel is design to hold three

membrane module inserts, but currently the module only holds two membrane module

inserts. The membrane module pressure vessels is a cylindrical tube, which is machined

on the inside to a fine tolerance. The roughness of the machined surface is of critical

importance with respect to the seal of the membrane modules via the U-cup seals. The

membrane has a 2” outlet nozzle for the permeate gas (end of vessel) and reside gas

(bottom of vessel) streams.

8

Table 1: Description of Major Parts Membrane Module Inserts: VMI-1, VMI-2 2x8” High-Pressure Membrane Module Inserts

Manufacturer: MTR

Serial No: 6665, 6666

Membrane Module Pressure Vessel: M-101 7.60” I.D. Membrane Module Pressure Vessel

Design Pressure: 1213 psig

Design Temperature: 150°F

MDMT: 0°F at 1213 psig

Material: Carbon Steel

Manufacturer: Progressive Recovery, Inc.

(Dupo, IL.)

Dwg. No:

AB71 – 8” 600# FLAT MEMBRANE

SECTION and

AB71 – 8” 600# FLAT VESSEL

ASSEMBLY

Codes: ASME Sect. VIII Div. 1

Filter Vessel: F-1 12” SCH80 Shell Filter Vessel


Design Temperature: 150 °F

MDMT: 0°F at 1213 psig

Material: Carbon Steel

Manufacturer: Progressive Recovery, Inc.

(Dupo, IL.)

Dwg. No:

AB71 – 8” 600# FLAT FILTER SECTION

and

AB71 – 8” 600# FLAT VESSEL

ASSEMBLY

Codes: ASME Sect. VIII Div. 1

Piping ASME B31.3


Design Temperature: 150 °F

Flanges ANSI CL 600

Manufacturer Process Connections

The Fuel Gas Conditioning Unit is designed as a complete unit. The following

process connections are required to make the unit operational:

1. Design and prepare anchoring locations and surface.

2. Allow enough space on the membrane module pressure vessel permeate head

side of the skid (at least 50 inches) to remove and re-install membrane module

inserts.

9

3. Anchor the skid to anchor locations.

4. Connect NOZZLE A to the feed gas line.

5. Connect NOZZLE B to the residue outlet line.

6. Connect NOZZLE C to the permeate outlet line.

7. Connect NOZZLE D to the liquid dump drain line.

8. Connect the flare/vent header to the PSV-193 blowdown line.

9. Perform a leak test with nitrogen prior to start-up according to normal plant

operating procedures.

It is very important to carry out leak testing in a way that will not damage the

module inserts. Never allow the permeate-side pressure to exceed the feed/residue- side

pressure. Never expose the membrane modules to liquid, and do not “shock” the modules

with a sudden pressure increase.

Manufacturer General Rules

The MTR FuelSep® membrane modules are robust and simple in construction.

However, they are vulnerable to damage in the following ways:

Permeate Over-Pressure

Never allow the permeate-side pressure to exceed the feed/residue side pressure.

The membrane module inserts are designed to be pressurized from the feed/residue side

only. Over-pressure or back-flow in the central permeate tube can cause physical damage

to the membranes. The membrane/spacer layers may separate, creating flow by-pass

channels. The net result will be reduced separation efficiency. Spring-loaded check

valves are permanently installed on the permeate tube lines to avoid such a scenario.

10

Liquid Contact

Never allow organic liquids to come into contact with the membrane module

inserts. Organic liquids will cause swelling of the membrane and spacers, thus reducing

feed flow and increasing the feed/residue differential pressure. Also, organic liquids may

be absorbed into the membrane itself, effectively “blinding” it.

Pressure Shock and Flow Surges

Sudden and dramatic changes in pressure and flow rates inside the module

pressure vessels should be avoided. Do not “shock” the modules with a sudden pressure

rise. Although prototype modules have survived pressure cycling tests without damage, it

is good practice to avoid pressure shocks. A pressure rise of 29-43.5 psi per minute is

acceptable. When starting up the system, the feed inlet control valve must control the gas

flow increase to the modules. The feed gas volume flow to the modules during start-up

should increase at the same rate as the pressure.

Maximum Differential Pressure Feed to Reside

The maximum differential pressure from the feed inlet port to the residue outlet

port of one membrane module pressure vessel with on membrane insert should not

exceed 5 psi.

Maximum Operating Temperature

The maximum operating temperature for the module inserts is 140°F. For optimal

performance, the inserts should not be operated at more than 100°F.

Oil Vapor

Oil vapor in concentrations of more than 1 ppm (wt) should not be introduced into

the membrane module inserts.

11

HCL

The maximum HCL content in the feed stream should not exceed 1 ppm (wt).

Instrumentation Guide

The FuelSep ® unit is held on a single skid. All controls are operated

pneumatically via instrument air. The instruments below in Table 2 are the variables to

consider.

Table 2: Instruments Used Tag. No. Description Function

XV-110 Control Valve

Opens when pressure on membrane residue outlet line reaches 350

psig, closes when under 350 psig. PV-110A senses pressure and

sends signal. SR-110C is a supply gas pressure regulator; it regulates

the instrument gas line down to a pressure of 20-35 psig. This setup

is located on the main feed line.

PSV-193 Pressure Relief

Valve

Set to relieve pressure on inlet gas line and protect the filter pressure

vessel F-1 and membrane pressure vessel M-101 when the pressure

exceeds 1213 psig. It is a Fire Safe type relief valve and is located

on the flare/vent line.

LS-593 Liquid Level

Switch

When liquid level rises enough within the filtration stage chamber of

the two-stage filter vessel and it is switched “on”, it sends a

pneumatic signal to the on/off liquid dump valve XV-511 to open up

and dump the liquids. It is located on the section of the bottle

chamber attached to the filtration stage of the two-stage filter vessel

F-1.

LS-594 Liquid Level

Switch

When liquid level rises enough within the knockout stage chamber

of the two-stage filter vessel and it is switched “on”, it sends a

pneumatic signal to the on/off liquid dump valve XV-512 to open up

and dump the liquids. It is located on the section of the bottle

chamber attached to the knockout stage of the two-stage filter vessel

F-1.

XV-511 On/Off Liquid

Dump Valve

It will open when the fluid accumulated within the filtration stage

chamber of the two-stage filter vessel F-1 rises to the level which

will switch the liquid level switch LS-593 on. It is located on the

filter vessel liquid dump line.

XV-512 On/Off Liquid

Dump Valve

It will open when the fluid accumulated within the knockout stage

chamber of the two-stage filter vessel F-1 rises to the level which

will switch the liquid level switch LS-594 on. It is located on the

filter vessel liquid dump line.

LG-590 Liquid Level

Gage

Indicates the fluid level accumulated within the filtration stage

chamber of the two-stage filter vessel F-1, and provides a visual

indication of the level. It is location on the section of the bottle

chamber attached to the filtration stage of the filter vessel F-1.

LG-591 Liquid Level

Gage

Indicates the fluid level accumulated within the knockout stage

chamber of the two-stage filter vessel F-1, and provides a visual

indication of the level. It is location on the section of the bottle

chamber attached to the knockout stage of the filter vessel F-1.

12

TI-291 Temperature

Indicating Gage

Indicates temperature on the membrane module pressure vessel

residue outlet line.

PI-292 Pressure

Indicating Gage

Indicates pressure on the membrane module pressure vessel residue

outlet line.

TI-391 Temperature

Indicating Gage

Indicates temperature on the membrane module pressure vessel

permeate outlet line.

PI-392 Pressure

Indicating Gage

Indicates pressure on the membrane module pressure vessel

permeate outlet line.

PI-595 Pressure

Indicating Gage

Indicates pressure on the combo (filter and membrane) vessel and is

located on the knockout stage of the vessel.

TI-597 Temperature

Indicating Gage

Indicates temperature on the combo (filter and membrane) vessel

and is location on the knockout stage of the vessel.

DPI-598

Differential

Pressure

Indicating Gage

Indicates the pressure difference between the knockout stage

chamber and the filtration stage chamber of the filter vessel section,

and has a pressure tap within each of these chambers.

Emerson

ROC

800

RTU

Series Remote

Operations

Controller

It is the device that collects all of the field data and transmits it to the

online data gathering system, Cygnet. It also has a digital display

screen that allows for the viewing of data in the field.

MVS

205

Multi-Variable

Sensor

Provides static pressure, differential pressure, and process

temperature inputs to a Remote Operations Controller (ROC).

Daniels

Simplex

Meter

Runs

Flowrate

Measurement

This is a standard 2” diameter meter run. It is the method in which

the flow rates are measured for the inlet, residue, and permeate

streams.

PROCEDURES

Preparation for Start-Up Procedure

1. All process pipes and electrical connections should be connected to the skid

properly.

2. Make sure all of the following manual valves are CLOSED:

a. Drain/Purge Valves:

i. Four x ¾” gate valves (V-160, V-272, V-372, V-572)

b. Sample Port Valves:

i. Three X ¾” gate valves (V-270, V-370, V-570)

c. MTR Skid Isolation Valves

i. One x 2” gate valve (V-700) on the main feed line @ TP-A

13

ii. One x 2” gate valve (V-200) on the membrane reside line @

TP-B

iii. One x 2” gate valve (V-300) on the rich gas line @ TP-C

iv. One x 1” gate valve (V-500) on the filter drain line @TP-D

3. Make sure all of the following manual valves are OPEN:

a. Instrument Gauge Valves:

i. Six x ½” gate valves (V-240, V-241, V-340, V-544, V-545, V-

546)

ii. Four x ¾” gate valves (V-540, V-541, V-542, V-543)

b. Isolation Valves:

i. One x 1” gate valve (V-143) in front of PSV-193 (lock open)

Start Up Procedure of FuelSep® System

1. Begin supplying the feed gas to the system through tie-point A, and slowly

open V-700. Slowly open the F-1 condensate drain outlet line isolation valve

V-500 at tie-point D. this will gradually pressurize the system and prevent any

shocking to the membrane module inserts.

2. When the system pressure reaches approximately 200 psig (PI-392) slowly

open the permeate purge line valve V-370.

3. Slowly open V-200 at tie-point B.

4. XV-110 should begin to open slowly as the pressure begins to build up

throughout the system and approach the “on” set pressure of 450 psig; PV-

110A is sensing the pressure on the reside line, via V-241. Once the pressure

14

throughout the system begins to equalize, the opening action of XV-110 will

level off.

5. Slowly open V-300 at nozzle C. Slowly close V-370.

6. Observe PI-292/PI-392 and TI-291/TI-391. Verify that all lines are at their

normal operating conditions.

Skid Purge Procedure

1. Begin supplying the purge gas to the system through tie-point A, and slowly

open V-700. This will gradually introduce the purge gas into the system and

prevent any shocking to the membrane module inserts.

2. Open V-370 simultaneously on the permeate line.

3. When the system pressure rises to 100 psig (PI-392), close V-700 to shut in

the supply of the purge gas to the system.

4. The system will be purged through V-370.

5. Slowly open V-270 to purge from the residue line.

6. Manually open dumb valves XV-511 and XV-512. Simultaneously open drain

valve V-572 on the liquid drain line.

7. System will slowly bleed through V-270, V-370, and V-572 and the pressure

will drop to atmospheric.

8. Close V-270, V-370, and V-572.

9. System is now purged and the pressure inside the system is now at

atmospheric pressure.

Shut Down Procedure of FuelSep® System

1. Slowly close the main feed isolation valve V-700 at tie-point A

15

2. Slowly close the membrane module pressure vessel residue outlet line

isolation valve V-200 at nozzle B.

3. At this point, there is no more incoming flow through the inlet at tie-point A,

and the remaining gas trapped within the membrane system will eventually be

forced to escape through the permeate line. Observe the pressure indicating

gauges PI-292, PI-392. The pressure should equalize throughout the system to

the permeate pressure of 20 psig.

4. Slowly close the membrane module pressure vessel permeate outlet (rich gas)

isolation valve V-300 at nozzle C.

5. To relieve the remaining pressure on the system, open the permeate purge

valve V-370. This will allow a blowdown of the remaining pressure without

causing any back-pressurization of the membrane module elements in the

vessel. Once all the pressure has been bled out of the system (as confirmed by

zero-readings on all pressure gauges) close V-370.

6. Close the F-1 condensate drain outlet line isolation valve V-500 at tie point D.

Phase I: Testing for Function of FuelSep® System

MTR provided a single-point operating condition of the MTR unit given the well

pad operating conditions and gas composition of the compressor discharge shown in

Table 3. The important streams to note are the inlet stream, the residue gas heading to the

engines fuel gas system, and the permeate gas heading to the suction header of the

compressor station. The Heat and Material Balance table below summarizes the estimated

system performance. Also, below is a diagram (Fig. 6) outlining the integration of the

MTR FuelSep® system with the compressor at the well pad facility.

16

Table 3: Operating Conditions at Well-Pad Facility Injection Total Flow 2400 MSCFD

Likely Total Max Flow Handling Capacity 3200 MSCFD (One 3508 Compressor)

Fuel Rate 90 MSCFD

Suction Pressure on Gas Lift 25 psig

Discharge Pressure on Gas Lift 700 psig

Sales Line Pressure 85 to 100 psig

Gas Composition Shown below

Table 4: Heat and Material Balance (Model Prediction)

Stream Name Inlet Residue (Fuel) Permeate (To

Gathering)

Stream No. 22 3 5

Overall

Molar Flow [lbmol/h] 30.1285 9.8823 20.2462

Mass flow [lb/h] 663.246 186.2934 476.9530

Temperature [°F] 100.8 74.5 87.7

Pressure [psia] 415 414 95

Vapor Mole Fraction 1.000 1.000 1.000

STD Specific Gravity 0.760 0.651 0.813

Heating Values (60°F)

Gross BTU/lbmol 4.354E+005 3.908E+005 4.572E+005

Net BTU/lbmol 3.947E+005 3.529E+005 4.151+005

Average Mol Wt. 22.0139 18.8513 23.5576

Actual Vol. [MMCFD] 0.0097 0.0031 0.0294

STD Liquid GPD 5118.8936 1574.6980 3544.1951

STD Vapor 60°F

MMSCFD 0.2744 0.0900

0.1844

Vapor Only

Molar Flow [lbmol/h] 30.1285 9.8823 20.2462

Mass Flow [lb/h] 663.2463 186.2934 476.9530

STD Vapor 60°F

MMSCFD 0.2744 0.0900

0.1844

Component Mole %

Carbon Dioxide 6.284190 3.189610 7.794665

Hydrogen Sulfide 0.000000 0.000000 0.000000

Nitrogen 1.502597 3.069078 0.737993

Methane 75.891155 86.247981 70.835942

Ethane 8.674885 4.943058 10.496403

17

Propane 4.737492 1.907245 6.118945

I-Butane 0.3706999 0.089459 0.507973

N-Butane 1.278298 0.308487 1.751667

I-Pentane 0.291999 0.063511 0.403525

N-Pentane 0.351299 0.071061 0.488085

N-Hexane 0.617199 0.110491 0.864524

N-Heptane 0.000000 0.000000 0.00000

Water 0.000189 0.000015 0.000273

Figure 6: Model of Membrane Performance at Well Pad Facility

The first phase in the performance testing of the MTR unit is to operate it at the

given optimal operating conditions and compare the G.C. of the residue and permeate

outlet lines to the model provided by MTR. The following procedure outlines how this

will be done. Note: the G.C. cycles every 5 minutes and always record time of operation

of the MTR unit.

1. Insure that XV-110 is set to open at 350 psig and regulate the following

pressure control valves: PR-710 to 430 psig, PR-200 set to lowest possible

18

pressure, roughly 70 psig, and the PR 310 to 100 psig. Record the fuel gas line

pressure and the raw gas line pressure. Follow the procedures outlined above

for initiating start-up and start-up of the unit.

2. Record the G.C., temperature (TI 597, TI 391, TI 291), and flow rates of the

residue, permeate, and inlet streams through Cygnet. Confirm that the

pressures are set to as specified in Step 1 (PI 595, PI 392, PI 292). Record the

fuel gas line pressure and the raw gas line pressure.

3. Compare the experimental G.C.s and flow rates to the model prediction.

4. Determine the performance index.

Phase II: Optimizing Performance Test of FuelSep® System

The purpose of the second phase of performance testing is to generate a

performance map that shows the ideal operation range of the MTR unit. Hopefully, the

performance map will demonstrate at what conditions the unit will meet the fuel

consumption rate of a given compressor and have the leanest residue composition

possible. Note: GC cycles every 5 minutes and always record time of operation of the

MTR unit.



pressure, roughly 70 psig, and PR 310 to 100 psig. Record the fuel gas line


for initiating start-up and start-up of the unit if unit is not on already. Allow

time for the unit to reach a steady state.

19

2. Record the initial G.C., temperature (TI 597, TI 391, TI 291), and flow rates

of the residue, permeate, and inlet gas streams through Cygnet. Confirm

pressures are set to as specified in Step 1 (PI 595, PI 392, PI 292).

3. Reset PR-710 to 475 psig located on the inlet gas stream. Record the G.C.,

flow rates, temperature (TI 597, TI 391, TI 291), and pressures (PI 595, PI

392, PI 292) of the residue, permeate, and inlet gas streams once a steady state

has been reached and the G.C. has cycled.

4. Repeat step 3 but vary the PR-710 setting in increments of 25 psig. Can

increase or decrease.

5. Determine the performance index for each variation in inlet pressure.

6. Reset PR-710 to the normal operating pressure of 430 psig.

7. Reset PR 310 to 95 psig located on the permeate gas stream. Record the G.C.,

flow rates, temperature (TI 597, TI 391, TI 291), and pressures (PI 595, PI

392, PI 292) of the residue, permeate, and inlet gas streams once a steady state

has been reached and the G.C. has cycled.

8. Repeat step 6 but vary the PR 310 setting in increments of 5 psig. Can

increase or decrease.

9. Determine the performance index for each variation in permeate pressure.

10. Record the fuel gas line pressure and the raw gas line pressure.

Phase III: Testing for Upsets of FuelSep® System

Once it is clear that the MTR unit is operational, it is important to understand how

it interacts with the system surrounding it when there are upsets. Ideally, the compressors

should not quit running in the event of a membrane malfunction. Instead, the compressors

20

should start pulling directly from the suction header gas line for fuel when the MTR quits

supplying residue gas. The following procedure will test this scenario and confirm no

effect on the compressor. Additionally, there needs to be an understanding of what

happens to the membrane itself if the residue outlet valve or permeate outlet valve is shut

during operation. The following procedure will test this scenario as well.



pressure, roughly 70 psig, and the PR 310 to 100 psig. Record the fuel gas line


for initiating start-up and start-up of the unit if unit is not on already. Allow

time for the unit to reach a steady state.

2. Record the initial G.C., temperature (TI 597, TI 391, TI 291), and flow rates

of the residue, permeate, and inlet gas streams through Cygnet. Confirm the


3. Slowly shut the V-300 valve located on the permeate gas line.


residue and inlet gas streams through Cygnet.

5. Compare the outcome to that outlined in the PHA.

6. If the MTR unit shut down, consult the start-up procedures to turn the unit on

again and allow it to come to a steady state before continuing through this

procedure.

7. If the unit did not shut down, slowly re-open V-300 and allow time for the

unit to reach steady state.

21

8. Record the initial G.C., temperature (TI 597, TI 391, TI 291), and flow rate of

the residue, permeate, and inlet gas streams through Cygnet. Confirm that the


9. Slowly shut the V-200 valve located on the residue gas line.


permeate and inlet gas streams through Cygnet. Record the fuel gas line

pressure and the raw gas line pressure.

11. Compare the outcome to that outlined in the PHA.

12. If the compressor shut down, follow the shutdown procedures of the MTR

Unit and call the foreman in order to restart the compressors.

13. If the compressor did not shut down, slowly reopen V-200 and allow the unit

to regain steady state.

14. Record the G.C. and flow rates of the residue, permeate, and inlet gas streams

through Cygnet.

Phase IV: Emissions Testing

Once an ideal point of operation of the membrane unit has been established, it

will be beneficial to see the effect of the fuel composition change on the compressor.

Therefore an emissions test will be performed on the compressor while operating with

rich fuel gas and while operating with the lean fuel gas produced by the membrane. The

emissions testing will need to be completed by an authorized person.

1. Set up an appointment with an authorized person for emissions testing.

2. Have the authorized person emission test the compressor for 20 minutes while

running with rich fuel.

22

3. Follow the start-up procedure in order to turn on the membrane to feed the

compressor lean gas. Shut off the feed of the rich fuel gas to the compressor.

This will insure that the compressor is solely running off of a lean gas supply.

The operating point of the membrane should be the point that has been proven

to provide the leanest supply of fuel gas.

4. Have the authorized person emission test the compressor for 20 minutes while

running with lean fuel.

5. Collect the data presented by both test and compare the results.

RESULTS

Phase I, Testing for Function, was conducted on the first day of operation, June

11th, 2014. The MTR unit was started by the procedures outlined by the manufacturer and

was set to operate at an inlet pressure of 415 psig, a permeate pressure of 70 psig, and a

fuel line pressure of 75 psig. The fuel regulator on the compressor skid was set for 60

psig. The MTR unit ran without issue, but the compressor did shut down because the fuel

line pressure was not high enough. No data was recorded on this day because it was just a

trial run to see how well the membrane system started and interacted with the compressor

before further testing was conducted. Phase II, Optimizing Performance, occurred over

several testing days and it outline below.

Table 5: Membrane Optimization Data

Test

Date

Inlet Pressure

Ranges [psig]

Fuel

Consumption

Rates [MCF]

Commentary

6/16/2014 360-510 73-95

XV-512 dumped 4 times off separator.

Condensation on line off of separator;

New spring on PR-200; Temperature

drop across PR-200 is 10 to 19°F

6/18/2014 384-472 94-99 Condensation line off of separator was

disconnected because it was leaking gas;

23

differential pressure and flow rates were

oscillating; XV-512 dumped twice off of

separator;

6/23/2014 385-540 110-281

Oscillations continued even after a tee off

fuel line was opened to allow for greater

fuel consumption rates; Stopped testing

early due to oscillations.

6/25/2014 407-701 100-245

Moved downstream sensor for PR-710 to

try and decrease oscillations. PR-710 still

oscillating in spurts. XV-512 dumped

three times; Temperature drop across PR-

200 was 24 to 35°F.

6/30/2014 364-665 99-218

PR-710 pilot valve replaced which

significantly reduced oscillations; XV-

512 dumped once; Temperature drop

across PR-200 was 28 to 34°F.

7/1/2014 525-615 144-168

Fuel consumption rate focused to roughly

150 MCF; Temperature drop across PR-

200 was 21 degrees; Fuel consumption

rate was not steady; Pressure change was

slow at 515 psig and fast at 580 psig.

7/7/2014 399-620 106-192

Fuel consumption rate was not steady;

Pressure change fast at 500 psig and slow

at 400 psig; No oscillations occurred at

585 psig and 600 psig.

Phase III, Testing for Upsets, occurred on June 14th, 2014 when the membrane

system was being prepared for overnight run time. While the membrane system as well as

the compressor were running, the fuel line to the compressor was closed. At that time the

PSV went off. Therefore, in the event that the fuel line is shut gas will be released into

the atmosphere uncontrollably. Unfortunately, the portion of testing that closed the

permeate line during operation was not conducted for fear of damaging the membrane.

Phase IV, Emissions Testing, was scheduled for June 23rd, 2014. The results from

that test are shown below.

24

Table 6: Emissions Testing Data

Component Percent Difference

Control vs. Test Gain or Reduction

%O2 6.54% Gain

ppm CO -21.52% Reduction

ppm NO -82.87% Reduction

ppm NO2 -43.03% Reduction

ppm NOx -60.70% Reduction

Figure 7: Nitrogen Oxide Emissions Reduction

Figure 8: Carbon Monoxide Emissions Reduction

25

Figure 9: Oxygen Emissions Reduction

Performance Index

The performance index is the means in which the data can be analyzed to

determine the optimal performance of the MTR unit as well as how sensitive it is to

changes in operating parameters. The performance index is a measure of membrane

efficiency and is based on the model predictions of the perfect division between permeate

and residue gas streams (i.e. the gas compositions) as well as the operating inlet and

outlet pressures. The ultimate goal is to have the permeate gas line as rich as possible and

the residue as lean as possible. When either one infiltrates the other, there is less

efficiency. Equation 1 is the way in which to calculate the performance index:

𝑃. 𝐼. = 1 − [𝑅𝑒𝑠𝑖𝑑𝑢𝑒𝐵𝑇𝑈 𝑀𝐶𝐹⁄ −1000

10∗ 0.09] − [

100−𝑅𝑒𝑠𝑖𝑑𝑢𝑒 𝑀𝑜𝑙 %

20∗ 0.15] (1)

26

Figure 10: Internal Rate of Return and Performance versus Performance Index

DISCUSSION

The most important analysis are the performance index and the Internal Rate of

Return, or IRR. Performance was calculated through consideration of the residue stream

using a sum of products method outline above while the IRR was calculated with the

permeate stream composition using an economics worksheet. The reason each of these

values are calculated differently is because on one hand this unit ought to be measured on

how well it produces a lean fuel going to the compressor while simultaneously producing

the richest stream possible to the sales gas gathering system. Therefore, the permeate

stream (rich gas) is what adds value and is used in economics because of the recovery of

natural gas liquids, or NGL’s, that are currently being burnt; and, performance is based

just on how lean the fuel is to the compressor. It is the combination of these two values

which will result in the best possible operating condition for the unit. Figure 10 shows

that as differential pressure increases so does performance and IRR. However, at some

27

point the IRR reaches its maximum and flat lines. In the median differential pressure

values there is a flat line as well for performance at various flow rates. The performance

then continues to increase, but only for high differentials with low fuel consumption

rates. Therefore, the best operating range is outlined by the red box.

Not only is the membrane system consistently giving a lower BTU/MCF content

for the compressor fuel, but also reducing the compressor’s emissions by 60% in NOx

and 21% in CO. This is a very important realization to this system because it is a viable

option to a compressor that has issues meeting environmental regulations. It is hard to

prove that the unit would be economical for the sole purpose of emissions, but it is clear

that there are intangible benefits to it. As a side note, this emissions test was done on a

compressor that is a lean burn system and did not have a catalyst. A catalyst is a material

that helps compressor reduce their emissions. It would be interesting to run an emissions

test on a compressor that is not a lean burn, and one that has a catalyst. The expectation

would be that the non-lean burn compressor would have a greater emission reduction that

the lean burn compressor, and an extended life period on the catalyst.

Assumptions Made

During testing, inevitably there will be assumptions made. For this particular

experiment, the assumptions include that no membrane degradation occurred while

testing, the unit had no inherent flaws, the inlet gas stream had a consistent gas

composition, there was a steady flow into the membrane inlet from the compressor,

instrumentation was calibrated and within tolerance, and all systems were set up correctly

as designed.

28

Economics

Economic analysis was completed on each data point. There are two angles of

approach when evaluating the economics of the membrane system, the first being fuel

savings. If the MTR unit were to run permanently, then the compressor will be running

off of leaner gas; therefore, the fuel cost charged to that well pad should be reduced

because the membrane is producing cheaper fuel gas than the fuel gas coming straight

from the well. Basically, the idea is that by having cheaper fuel costs, the lease operating

expense, LOE, should be reduced. Unfortunately, this does not show much promise.

From the economics analysis, the fuel savings would really only be roughly $12,000 a

year assuming a fuel consumption of 100MCF/day. This will not cover the cost of the

unit. Additionally, accounting allocates fuel on a MCF basis.

The second angle would be adding value back to the well because the membrane

is recovering NGL’s that would have normally been burnt by the compressor as fuel. By

tracking how much the membrane is recovering and the composition of that stream, there

would be a way to allocate the NGL’s back to the well pad site. By adding more

produced value to the well this should also reduce LOE. The economic spreadsheet used

for determining IRR is the view point of earning revenue through NGL recovery. Net

membrane revenue includes transportation and fractionation cost, a rough estimate of

membrane operation and maintenance cost, cost of compression as well as natural gas

plant recovery rates. If the gas were going to plant not company owned, then additional

reductions would need to be made in the net membrane revenue calculation. Any sort of

royalties were not accounted for, as well as membrane degradation over time before it

were to be replaced. The overall rate of return was greater than 100%.

29

Qualitative Benefits

Qualitative benefits include less emissions, longer lasting catalyst, less deposits

on valves for the compressor, which combine to decrease compressor maintenance cost.

More natural gas liquids maybe recovered at the plant if more units were implemented.

Also, head life on the compressor will be longer due to less fatigue due to burning a rich

fuel.

Plans for Long Term Installation

Before the membrane unit would be left for unattended, long term operation, a

couple of safety features would need to be added to increase user confidence. The first

would be a high differential pressure switch on the coalescing filter. In the event that the

coalescing filter reaches a high differential pressure and no longer is doing its job, there

would need to be a way to shut down the membrane’s operation to insure that the

membrane modules downstream of the coalescing filter would not be contaminated. The

membrane modules cost much more to replace than just the coalescing filter. Also, the

separator section of the membrane would need to have a high level shut down switch on

it as well. The liquids that ultimately build up in the separator section of the membrane

system would need to be prevented from reaching the membrane modules, so a high level

switch will make sure that will not happen. In addition to safety features, if the membrane

unit were to operate under normal conditions it would not have all of the instrumentation

that is currently on it. The membrane would probably only need to have a fuel meter.

Therefore, there would need to be a quarterly gas analysis that would confirm the unit is

operating as expected, and to monitor the water content in the fuel stream. Any kind of

gas stream with too much water could have freezing problems during the winter. In the

30

event that it seems economical to put a feedback loop on the membrane, it could be done

by actuating the inlet regulator to match the fuel consumption rate. Finally, the allocation

back to the well would need to be evaluated and the economics of the unit would need to

be refined once maintenance and operation cost of the membrane are better understood.

CONCLUSION

The MTR FuelSep® System effectively creates a lean fuel to the compressor. It

proves to be economical and functional at a variety of differential pressures and fuel

consumption rates. The empirically based performance map for the unit gives an idea as

to what differential pressure will result in the fuel consumption rate needed to maximize

performance. Additionally, the membrane (because it is providing a better fuel) is

reducing the compressor emissions and extending the head life by leaving less deposits

on the valves and fittings. Overall, this unit seems to do its job and do it well. This unit

would be further beneficial to areas such as San Antonio and Midland where the fuel gas

being burnt has a much higher concentration in heavier hydrocarbons than was seen in

this test.

31

REFERENCES

Alvarado, Dean, and Douglas E. Gottschlich. "Membrane Matrimony." Hydrocarbon

Engineering (2005): n. pag. Membrane Technology and Research Inc. Web.

Jacobs, Marc, and Kaaeid A. Lokhandwala. "New Membrane Applications in Gas

Processing." Paper presented at Gas Processors Association Annual Meeting,

Spring 2000. n. pag. Membrane Technology and Research, Inc. Web.

Jacobs, Marc, and Kaaeid A. Lokhandwala. “Membranes for Fuel Gas Conditioning.”

Hydrocarbon Engineering (2000). n. pag. Membrane Technology and Research

Inc. Web.

Markiewicz, G.S., M.S. Losin, and K.M. Campbell. “The Membrane Alternative for

Natural Gas Treating: Two Case Studies,” (1988). Air Products & Chemicals Inc.

Society of Petroleum Engineers. n. pag. OnePetro. Web.

Membrane Technology and Research, Inc. Web.

Myers, Richard O. “An Overview of Today’s Membranes and Membrane Processes,”

(2000). DuPont Permasep® Products. NACE International. n. pag. OnePetro.

Web.

Ohlrogge, Klaus, Jan Wind, and Torsten Brinkmann. “Membrane Technology for Natural

Gas Processing,” (2002). Society of Petroleum Engineers. n. pag. OnePetro.

doi:10.2118/75505-MS. Web.

Documents

MEMBRANE TECHNOLOGY: A METHOD OF GAS …